INDEX

Explanations

locations or places containing the word "San"

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

references to the location San or associated places

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 cover

-0.65

 subject

-0.65

 impart

-0.65

 externally

-0.65

 greens

-0.63

 bargaining

-0.63

 negotiating

-0.62

 markup

-0.62

 internally

-0.61

 professionally

-0.61

POSITIVE LOGITS

gha

1.21

ctions

1.19

Disk

1.19

 Francisco

1.18

jay

1.16

ofi

1.13

jee

1.13

rio

1.12

ctuary

1.11

 Marino

1.07

Activations Density 0.016%

locations or places containing the word "San"

references to the location San or associated places

No Comments

No Known Activations